My problem is similar to that of McDonald
(http://www.stata.com/statalist/archive/2003-05/msg00650.html).
I am using Stata 9 to run a Heckman selection model. If I estimate for
the whole sample, everything seems fine. But, if I split the sample it
results the message "initial values not feasible" r(1400). The error
does not occure, too, if I apply the "twostep" option. After that, I
tried to estimate the model by saving the coefficients of the
"twostep"-model [matrix b=e(b)] and running the simultaneous model with
the "from"-option [from(b, copy)] to set the initial values, but it
doesn't work nevertheless. Has anyone a tip what's wrong with the
specification?
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
